Jet Bank: Credit Risk Foundation for Albania’s First Digital Bank
Adastra built the entire credit risk foundation for Albania’s first fully digital bank: hands-on, from scratch, with no historical data, and a roadmap designed to grow with the bank.
first fully digital bank in the country and a pioneer in the region
months to build credit risk foundation from a blank page to production-ready
estimated time, in minutes, from loan application to money in account
Challenge
Building a Lending Engine for a Bank That Had Never Made a Loan
In Albania, banking has always meant branch visits, paper forms, and days of waiting. Jet Bank came with a bold vision to replace all of that: a customer applies through an app, gets a decision in seconds, and has money in their account within roughly 20 minutes.
That experience only works with a robust credit risk engine underneath it. But there was no internal data to build from. As a brand-new bank, Jet Bank’s team had to design a modern system from scratch, tailored to the Albanian market.
A significant portion of the working population in Albania is self-employed, which limits what data is available and complicates the development of models and processes. Also, as a country actively pursuing EU accession, with negotiations targeting membership by around 2030 under Albania’s Digital Agenda 2022-2026, Jet Bank needed to be fully compliant from day one.
A functioning framework had to be in place from launch to:
- Make every lending decision digital and fast, to deliver on the digital-first customer promise
- Minimize fraud exposure, since digital lending without face-to-face interaction is a natural target for high-risk applications
- Ensure compliance with IFRS 9’s provisioning obligations
- Manage risk with a calibrated scorecard to ensure portfolio quality
Solution
A Greenfield Credit Risk Foundation, Delivered in Six Months
Typically, building a credit risk framework of this scope takes a year or more. Adastra completed it in less than six months, working as an embedded partner with Jet Bank’s team on the ground.
The approach drew on our experience across 100+ financial institutions, including major regional banks, to construct rigorous models without being data-dependent. Government-accessible data (such as social contribution records, which establish employment history and approximate income), credit bureau information, and regional market intelligence replaced the internal portfolio history that the bank did not yet have.
Probability of Default and Fraud Estimation
The Probability of Default (PD) scorecard assesses credit risk and determines whether an application is approved, referred for further assessment, or declined, drawing on demographic and employment data, employment records, and credit bureau data accessed entirely through digital channels. Anti-fraud rules and early fraud indicators were built in from the start because, in a fully digital, no-branch environment, fraud is a day-one operational reality rather than a future concern.
Income Estimation
Adastra built a statistical income model using government social contribution data and credit bureau signals. Borrowing capacity can now be assessed accurately and digitally, without requiring customers to submit any physical documentation.
Debt Collections
The full collections operating model was designed before the bank had a single overdue account: customer segmentation, contact strategy, KPIs, organizational structure, and vendor selection for the collections management system. By the time the first loans went past due, everything was already in place.
IFRS 9 Compliance
A full implementation blueprint covering all key risk parameters: probability of default (PD), loss given default (LGD), exposure at default (EAD), and expected credit loss (ECL), with macroeconomic adjustments for the Albanian context, gave Jet Bank an auditable compliance framework from launch and supported its alignment with EU financial sector standards.
Beyond the four workstreams, Adastra delivered a roadmap: which data Jet Bank should collect from its own customers, how to use it to refine the models over time, and what the path to a fully analytics-driven credit function looks like as the portfolio grows. The foundation built today is the starting point. The roadmap enables Jet Bank to turn its data into a long-term competitive advantage.
Impact
Open for Business, Ready to Scale
Jet Bank launched on schedule, backed by a complete, implementation-ready credit risk framework. When fully deployed, the system is designed to deliver capabilities that most Albanian banks, which have been operating for decades, still lack.
- ~20 minutes from application to disbursement for approved loans without human interaction
- Instant credit decisions at any volume, from any channel
- IFRS 9 compliance in place before the first loan was booked
- Collections function designed and ready even before an account went overdue
- Analytics foundation and data roadmap in place so models improve as the portfolio grows
Jet Bank set out to prove that modern, responsible banking in Albania does not require a branch network, decades of legacy data, or years of preparation. With Adastra as a partner, the complete credit risk framework was designed, validated, and handed over in six months, built to last, and ready to scale.
